Section 01

Spotting Behavioral Patterns Without the Manual Grind

You know how it is. You sit through a dozen call recordings trying to figure out why a deal didn't close, and you're just hunting for needles in a haystack. You're scanning for that one moment where the rep interrupted the client or skipped a discovery question. It's draining, and honestly, you're missing the actual patterns. You're too close to the noise to see the behavior that's really holding the deal back.

Claude isn't just giving you a recap here. It acts like a fresh pair of eyes that doesn't get tired. Feed it a transcript and it analyzes the structure of the conversation. Did the rep dominate the talk? Did they miss a clear buying signal? It flags the exact spots where momentum died. You get an objective breakdown of how the call actually went, not just a summary of what was said.

Key insight: Claude performs structural pattern matching, meaning it identifies behavioral gaps — like excessive talk time or missed discovery phases — that aren't obvious in a standard narrative summary.

Sample Call Coaching Scorecard

Talk-to-listen ratioRep spoke 78% of the call

Needs Work

Discovery questions askedOnly 2 open questions in 30 minutes

Missed

Buying signal responseClient mentioned budget — rep moved on

Missed

Objection handlingAcknowledged concern before pivoting

Strong

Next step commitmentClear follow-up date agreed on call

Strong

Behavioral isolation

Pinpoint specific moments where a rep talked over a client or missed a chance to dig deeper.

Objection assessment

Review how well a rep handled pushback and compare it against the best way to handle that same objection.

Talk-time analysis

Break down the balance of the conversation to see if the rep is asking enough questions or just pitching straight.

Without Claude

You spend your whole week listening to recordings, trying to remember every mistake you heard, and your feedback ends up being vague and unhelpful.

With Claude

You get a breakdown of the structural flaws in the call, so your coaching is focused, fair, and based on what actually happened.

You don't need a formal training program to do this. Just follow this rhythm to get your reps the feedback they need.

Record the call

Use your usual platform so you have the raw audio ready to process.

Grab the transcript

Download the transcript file once your platform has finished processing the recording.

Define your coaching criteria

Before opening the AI, decide what you're looking for — listen-to-talk ratio, objection handling, discovery depth.

Paste into Claude

Feed the transcript and your criteria into Claude so it knows exactly what to look for.

Generate the critique

Read through the AI's structural breakdown and note the specific moments it flags.

Debrief with your rep

Take the findings into a one-on-one and talk through what happened — together, not at them.

Note: Claude can't hear your rep's tone or hesitation in their voice, so don't treat the critique as the final word on their soft skills.

Don't just ask Claude to "look at this." If you're lazy with your prompt, you'll get lazy feedback. Use these to get the real insight.

Prompt 1 — Structural behavioral critique

Act as a sales coach. Analyze this transcript for structural issues: [Insert Transcript]. Specifically, tell me where the rep talked too much, where they missed a discovery question, and where they failed to address a buyer concern. Crucial instruction: Provide timestamps or specific quotes for every critique. Do not generalize; be specific about the behavior.

Prompt 2 — Objection handling review

Look at how the rep handled this specific objection: [Insert Objection]. Compare it to our internal playbook approach: [Insert Playbook]. Identify the exact moment the rep lost control of the conversation. Crucial instruction: Keep your feedback direct. Don't fluff it up. Tell me what they should have said instead.

Before taking this critique into a one-on-one

Specificity: are your points based on actual quotes, or just your general feeling?

Balance: did you also highlight what the rep did right, or is it just a list of mistakes?

Actionability: can the rep actually fix the problem by changing one specific behavior next time?

Context: did you account for the specific client personality in the call, or are you judging the rep in a vacuum?

Tone: does your feedback encourage the rep, or does it sound like a robot listing errors?

Important: Use AI feedback as a starting point, not a replacement for human manager judgment. You know the history of the rep and the relationship with the client — the AI doesn't. Always filter the output through your own experience before giving it to your team.